Changes In Treatment Patterns And Overall Survival For Early-Stage Diffuse Large B Cell Lymphoma In The United States After Rituximab: A Population-Based Analysis

The introduction of Rituximab for first line treatment for diffuse large B-cell lymphoma (DLBCL) in 2002 has improved oncologic outcomes and led to further questions regarding the benefit of consolidative RT for DLBCL. This study sought to determine changes in RT utilization for the treatment of Stage I-II DLBCL after the incorporation of Rituximab as first line treatment of DLBCL. This population-based study also compared survival outcomes for patients treated with or without RT in both the pre-Rituximab era (1992-2001) and post-Rituximab era (2002-2011) and aimed to identify factors that affected outcome for these patients. We included patients in the Surveillance, Epidemiology, and End Results (SEER) database, diagnosed with Stage I-II DLBCL between 1992 and 2001. Changes in treatment patterns were assessed using chi-squared test. Simple linear regression was performed to determine the rate of RT utilization over time as a function of year of DLBCL diagnosis during the pre-Rituximab era and post-Rituximab eras. Kaplan-Meier and Cox Regression were performed to compare overall survival (OS) for patients treated with or without RT during these time periods. Multivariate analysis was performed to identify demographic, treatment, and tumor characteristic predictive of OS. A total of 38,407 patients met the specified criteria, of whom 33% received radiation. Median follow-up time was 39 months (interquartile range, 10-91 months). RT utilization was 37.7% in the pre-Rituximab era and 30.6% in the post-Rituximab era (P < 0.001). Simple linear regression revealed that in the pre-rituximab era the slope of the best fit line for RT utilization by year was positive (m = 0.006, P = 0.0233) and the slope for the best fit line in the post-rituximab era was negative (m = -0.009, P = 0.0029). In the pre-Rituximab era, RT use was associated with significant improvements in OS, with a 5-yr OS for patients treated with RT of 57% compared to 5-yr OS without RT of 45.2% (hazard ratio [HR] = 0.746; 95% confidence interval 0.714 to 0.778). In the post- Rituximab era, RT was associated with significant improvements in OS, with a 5-yr OS for patients treated with RT of 68.3% with RT compared to 5-yr OS of 53.4% without RT (HR = 0.588; 95% confidence interval = 0.563 to 0.615). Matched pair analyses confirmed that RT use improved OS in the pre-Rituximab era (HR = 0.815; 95% confidence interval = 0.777 to 0.856) and in the post-Rituximab era (0.692; 85% confidence interval = 0.657 to 0.728). On multivariate analysis, RT was associated with improved OS. There has been a decrease in RT utilization following the incorporation of Rituximab for first line treatment of DLBCL in the United States. Use of RT for management of Stage I and II DLBCL is associated with improved OS in both the pre- and post- Rituximab era. The present report suggests that RT should be used for management of early stage DLBCL due to improved survival outcomes.